Loic Dachary
0fb0967e95
documentation: tiering typo
...
Signed-off-by: Loic Dachary <loic-201408@dachary.org>
2014-09-19 15:05:27 +02:00
Gregory Farnum
9442336f92
Merge pull request #2514 from ceph/wip-acl-test
...
qa/workunits/fs/misc: Add a workunit for ACL
Reviewed-by: Greg Farnum <greg@inktank.com>
2014-09-18 15:01:15 -07:00
Samuel Just
5c8143e3d5
Merge pull request #2457 from ceph/wip-7767-b
...
msg: prefetch data when doing recv()
Reviewed-by: Greg Farnum <greg@inktank.com>
Reviewed-by: Samuel Just <sam.just@inktnak.com>
2014-09-18 09:28:05 -07:00
Samuel Just
a8e6de307f
Merge pull request #2440 from somnathr/wip-optracker-optimization
...
Wip optracker optimization
Reviewed-by: Samuel Just <sam.just@inktank.com>
2014-09-18 09:26:40 -07:00
Samuel Just
95768963f2
Merge pull request #2445 from wonzhq/req-pending
...
Cache tiering: do not allow flush/evict when there are requests pending
Reviewed-by: Samuel Just <sam.just@inktank.com>
2014-09-18 09:25:18 -07:00
John Spray
ab545ad0f9
Merge pull request #2515 from ceph/wip-9518
...
mds: fix not journaling client metadata
Reviewed-by: Yan, Zheng <ukernel@gmail.com>
2014-09-18 13:53:45 +01:00
John Spray
1395275e0f
mds: fix not journaling client metadata
...
Previously the code was there for storing in
the SessionMap table, but not for the ESession
logevent.
Fixes : #9518
Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-18 10:29:06 +01:00
Yan, Zheng
59c89760c7
qa/workunits/fs/misc: Add a workunit for ACL
...
Signed-off-by: Yan, Zheng <zyan@redhat.com>
2014-09-18 14:20:59 +08:00
Yan, Zheng
051cc7d525
Merge pull request #2503 from ceph/wip-xattr-fix
...
Wip xattr fix
2014-09-18 10:55:43 +08:00
Josh Durgin
829981792a
Merge remote-tracking branch 'origin/giant'
2014-09-17 12:23:17 -07:00
John Spray
cef34f4299
Merge pull request #2498 from ceph/wip-objecter-cancel
...
osdc/Objecter: cancel timeout before clearing op->session
Reviewed-by: John Spray <john.spray@redhat.com>
2014-09-17 19:07:22 +01:00
Gregory Farnum
f27ccd805e
Merge pull request #2513 from ceph/wip-decode-dump
...
msg: dump messages we can' decode at a low devel level
Reviewed-by: Greg Farnum <greg@inktank.com>
2014-09-17 10:15:03 -07:00
Sage Weil
4089684259
msg: hexdump bad messages at debug level 1 (not 30)
...
These are rare, and when they do happen, we want to see them in the
logs.
Signed-off-by: Sage Weil <sage@redhat.com>
2014-09-17 09:53:36 -07:00
Sage Weil
54955702fc
msg: hexdump messages we can't decode at configurable debug level
...
We had this hard-coded to 30. Make it tunable.
Signed-off-by: Sage Weil <sage@redhat.com>
2014-09-17 09:52:54 -07:00
John Spray
d8630cf446
Merge pull request #2512 from dachary/wip-9490-crushtool-num-rep
...
Revert "crushtool: safeguard for missing --num-rep when --test"
Reviewed-by: John Spray <john.spray@redhat.com>
2014-09-17 13:59:33 +01:00
Loic Dachary
480c3727c9
Revert "crushtool: safeguard for missing --num-rep when --test"
...
This reverts commit 10b8966c8d
.
2014-09-17 14:56:10 +02:00
John Spray
51eeb3340e
Merge pull request #2496 from ceph/wip-9189
...
#9375 and #9189 - add client metadata and improve health messages
Reviewed-by: Greg Farnum greg@inktank.com
2014-09-17 13:47:27 +01:00
John Spray
288f05a58f
mds: use new Session::human_name in health msgs
...
Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-17 13:21:49 +01:00
John Spray
6320e53569
mds: implement const SessionMap::get_session
...
Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-17 13:21:49 +01:00
John Spray
32b51bb7d5
mds: generate friendly name for client sessions
...
...from the client metadata, for use in health messages.
Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-17 13:21:49 +01:00
John Spray
920cac907b
ceph_fuse: include mount point in client metadata
...
Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-17 13:21:48 +01:00
John Spray
bd8c441351
client: send metadata in session open
...
Populated with hostname and entity_id by
default, with interface for outer layers
like ceph_fuse, libcephfs to inject their
own metadata.
Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-17 13:21:48 +01:00
John Spray
bdda237751
ceph_fuse: refactor init to make mount_point accessible
...
...so that we can read out the mount point between init
and start() in order to feed it to Client before it connects
to the MDS.
Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-17 13:21:48 +01:00
John Spray
758a7ceaf7
mds: include client metadata in session ls
...
Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-17 11:26:41 +01:00
John Spray
1a9e3cc99a
mds: store client metadata on Session
...
Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-17 11:26:41 +01:00
John Spray
cd215c7561
messages: add cephfs client metadata to MClientSession
...
Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-17 11:26:33 +01:00
John Spray
3025b5f09a
mds: collapse numerous client warnings into one
...
...to avoid sending O(N_clients) sized beacons.
Fixes : #9375
Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-17 11:25:03 +01:00
Loic Dachary
66deda4b45
Merge pull request #2508 from dachary/wip-9490-crushtool-num-rep
...
crushtool: safeguard for missing --num-rep when --test
Reviewed-by: Johnu George <johnugeorge109@gmail.com>
2014-09-17 09:06:44 +02:00
Yan, Zheng
1f450faaac
client: request xattrs if xattr_version is 0
...
Following sequence of events can happen.
- Client releases an inode, queues cap release message.
- A 'lookup' reply brings the same inode back, but the reply
doesn't contain xattrs because MDS didn't receive the cap release
message and thought client already has up-to-data xattrs.
The fix is force sending a getattr request to MDS if xattrs_version is 0.
The getattr mask is set to CEPH_STAT_CAP_XATTR, so MDS knows client does
not have xattr.
Signed-off-by: Yan, Zheng <zyan@redhat.com>
2014-09-17 11:32:17 +08:00
Yan, Zheng
c837fb9428
mds: return xattrs when client explictly requests xattrs
...
Signed-off-by: Yan, Zheng <zyan@redhat.com>
2014-09-17 09:38:27 +08:00
Gregory Farnum
eb1b2e0072
Merge pull request #2507 from ceph/wip-9435
...
#9435 - enforce rules for using EC/tiered pools with CephFS
Reviewed-by: Greg Farnum <greg@inktank.com>
2014-09-16 17:37:26 -07:00
John Spray
681a49c412
mon: forbid tier changes when in use by FS
...
* Removing tiers from a base pool in use by CephFS is forbidden.
* Using CephFS pools as tiers is forbidden.
Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-16 17:16:23 -07:00
John Spray
80441cda8c
mon: prevent cache pools being used CephFS
...
Fixes two things:
* EC pools are now permissible if they have a cache overlay
* Pools are not permissible if they are a cache tier.
Fixes : #9435
Signed-off-by: John Spray <john.spray@redhat.com>
2014-09-16 17:16:23 -07:00
Loic Dachary
10b8966c8d
crushtool: safeguard for missing --num-rep when --test
...
http://tracker.ceph.com/issues/9490 Fixes : #9490
Signed-off-by: Loic Dachary <loic-201408@dachary.org>
2014-09-16 21:18:52 +02:00
Josh Durgin
ea48a21ec9
Merge pull request #2475 from ceph/wip-lttng
...
Update rbd-replay
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
2014-09-16 12:15:30 -07:00
John Wilkins
13b82ded62
Merge pull request #2473 from ceph/wip-doc-preflight
...
doc: Revised the Preflight Doc
Reviewed-by: John Wilkins <jowilkin@redhat.com>
2014-09-16 11:47:12 -07:00
John Wilkins
9a65b39022
doc: Fixed broken hyperlink.
...
Signed-off-by: John Wilkins <jowilki@redhat.com>
2014-09-16 11:28:19 -07:00
Sage Weil
399fa80625
script/run-coverity: update submodules
...
Signed-off-by: Sage Weil <sage@redhat.com>
2014-09-16 11:01:29 -07:00
Loic Dachary
cf17111ceb
Merge pull request #2394 from ceph/wip-da-fix-make_check
...
Fix some issues with make check
Reviewed-by: Loic Dachary <loic-201408@dachary.org>
2014-09-16 15:26:40 +02:00
Loic Dachary
52c5c32f2c
Merge pull request #2504 from dachary/wip-erasure-code-valgrind
...
test: libtool --mode=execute valgrind
Reviewed-by: Sahid Orentino Ferdjaoui <sahid.ferdjaoui@redhat.com>
2014-09-16 15:22:10 +02:00
Loic Dachary
fdbfece81c
Merge pull request #2497 from ceph/wip-xfs-inode64
...
ceph-disk: mount xfs with inode64 by default
Reviewed-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
2014-09-16 15:12:24 +02:00
Danny Al-Gaaf
123c3fbbb7
test/mon/mkfs.sh: add check for default keyring
...
The auth_cephx_key check always fails if there is a default
keyring in /etc/ceph/ available. Check if keyring exists and
fail with error message if keyring file in path.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
2014-09-16 12:00:08 +02:00
Danny Al-Gaaf
8ea86dfa7c
test/ceph-disk.sh: resolve symlinks before check
...
Make sure symlinks are resolved in command_fixture()
before compare result of which command and the current
path.
Signed-off-by: Danny Al-Gaaf <danny.al-gaaf@bisect.de>
2014-09-16 12:00:04 +02:00
Loic Dachary
f952f45307
Merge pull request #2487 from wonzhq/org
...
mailmap: correcting Zhiqiang Wang's mailmap and org affiliation
Reviewed-by: Loic Dachary <loic-201408@dachary.org>
2014-09-16 10:55:13 +02:00
Zhiqiang Wang
daf63798dd
mailmap: correcting Zhiqiang Wang's mailmap and org affiliation
...
Signed-off-by: Zhiqiang Wang <wonzhq@hotmail.com>
2014-09-16 16:46:07 +08:00
Loic Dachary
6d4e2ca02c
test: libtool --mode=execute valgrind
...
otherwise valgrind will not actually run against the binary.
Signed-off-by: Loic Dachary <loic-201408@dachary.org>
2014-09-16 10:25:40 +02:00
Yan, Zheng
306fb2f5e9
mds: set new inode's xattr version to 1
...
set new inode's xattr version to 1 even if it has no xattr. This allow
client to differentiate no xattr in inode from MDS skips sending xattr
to client (because MDS think client already has uptodate xattr).
Signed-off-by: Yan, Zheng <zyan@redhat.com>
2014-09-16 12:54:48 +08:00
Sage Weil
782848af59
Merge pull request #2499 from ceph/wip-9219-giant
...
wip-9219: subscribe to the newest osdmap when reconnecting to a monitor
Reviewed-by: Sage Weil <sage@redhat.com>
2014-09-15 17:40:28 -07:00
Greg Farnum
1b9226c723
osd: subscribe to the newest osdmap when reconnecting to a monitor
...
This is mostly relevant in testing clusters, but it ensures that an OSD
disconnecting from the monitor at the wrong time will still see any recent
map updates and prevent accidental loss of map injection into the OSD cluster.
Fixes : #9219
Signed-off-by: Greg Farnum <greg@inktank.com>
2014-09-15 17:07:41 -07:00
Sage Weil
56ba341174
osdc/Objecter: fix command op cancellation race
...
Cancel the command op timeout event before we clear out the op from the
session struct. This isn't strictly necessary because command_op_cancel
will "gracefully" handle the case where the tid is no longer present, but
this avoids that noise and is cleaner.
Signed-off-by: Sage Weil <sage@redhat.com>
2014-09-15 16:45:19 -07:00